Re: Use LIMIT instead of Unique for DISTINCT when all distinct pathkeys are redundant

When all the distinct pathkeys are redundant then there can only be,
at most, 1 single distinct value. There may be many rows with that
value, but we can remove those extra ones with a LIMIT 1 rather than
troubling over needlessly uniquifing them.

This might not be a hugely common case, but; 1) it is very cheap to
detect and 2) the speedups are likely to be *very* good.

With the attached we get:

regression=# explain (analyze, costs off, timing off) SELECT DISTINCT
four,1,2,3 FROM tenk1 WHERE four = 0;
                   QUERY PLAN
-------------------------------------------------
 Limit (actual rows=1 loops=1)
   ->  Seq Scan on tenk1 (actual rows=1 loops=1)
         Filter: (four = 0)
 Planning Time: 0.215 ms
 Execution Time: 0.071 ms

naturally, if we removed the WHERE four = 0, we can't optimise this
plan using this method.

I see no reason why this also can't work for DISTINCT ON too.